Documentation & System Guides
Authoritative technical specifications and developer guides for the FineTuneMyAI local-first platform.
Quickstart & Installation
Install the native daemon on Linux, Windows, or macOS and connect your workstation.
Device Pairing Protocol
Understand ephemeral 6-digit one-time tokens, public-key authentication, and device revocation.
Local Model Registry
How the agent maps opaque identifiers (model_*) to on-disk weights without transmitting paths.
Corpus Quality & Preprocessing
Evaluating duplicate rates, boilerplate ratios, and token length distributions.
Fine-Tuning Hyperparameters
LoRA rank (r), alpha (α), sequence length, batching, and QLoRA 4-bit calibration.
Evaluation & Benchmarking
Side-by-side inference, perplexity tracking, validation loss, and latency benchmarking.
Local RAG Architecture
Semantic chunking presets, candidate top-k vs rerank k, and retrieval evaluation.
VPS & Production Server Deployment
Deploy the sovereign control plane on Ubuntu/Debian VPS using Docker Compose, Systemd, and Nginx SSL.
CyberPanel & OpenLiteSpeed Deployment
Native non-Docker deployment on CyberPanel VPS with built-in MariaDB (port 3306), PM2, and OpenLiteSpeed.